wade japan

JAPAN, THE WORLD BANK, AND THE ART OF PARADIGM MAINTENANCE: THE EAST ASIAN MIRACLE IN POLITICAL PERSPECTIVE To what extent is the World Bank an actor, an ‘autonomous variable’ in the international system?footnote1 Or to what extent are its objectives and approaches the mere manifestations of competition and compromise among its member states? Several writers have argued that the Bank has a relatively large amount of autonomy—from the state interests of its overseers, and that its staff have some autonomy from the senior management. They have traced this autonomy to variables such as ‘lack of clarity of the priorities of organizational objectives’, ‘the difficulty and complexity of accomplishing the organization’s mandate’, ‘bureaucratized structure’ and ‘professionalism of staff’.footnote2 But there is something strangely bloodless about this approach. It manages to discuss autonomy without conveying anything of the political and economic substance of the field of forces in which the Bank operates. By focusing only on morphological variables like ‘professionalism’ and the ‘complexity of accomplishing the organization’s mandate’, it misses other variables like ‘correspondence of organizational actions with the interests of the us state’. If the Bank is propelled by its budgetary, staffing and incentive structures to act in line with those interests, the us state need not intervene in ways that would provide evidence of ‘lack of autonomy’; yet the Bank’s autonomy is clearly questionable. This paper describes an episode in Japan’s attempts over the 1980s and 1990s to assert itself on the world stage, to move beyond the constraints of dependency in a us-centred world economic system. The episode involves a Japanese challenge to the World Bank and its core ideas about the role of the state in the strategy for economic development. Over the 1980s Japan poured aid and investment into East and Southeast Asia, using its strong domestic capacity to strengthen its external reach. In doing so, Japan endorsed a market-guiding role for the state in recipient countries, and justified this role by pointing to its success in the development of Japan, Taiwan, and South Korea. The World Bank found Japan’s prescriptions inconsistent with its own programmatic ideas about the role of the state, which emphasized the need for thoroughgoing liberalization and privatization. Since the Bank’s ideas are themselves derived from largely American interests in and ideas about free markets, Japan’s challenge to the Bank was also a challenge to the us state—the Bank being an important instrument by which the us state seeks to project a powerful external reach, while having a much weaker domestic capacity than Japan’s. In the early 1980s, when the Bank started to champion liberalization and the private sector, the Bank and the Japanese government proceeded along independent paths. But growing tension reached a head in the late 1980s when the Bank criticized Japanese aid programmes, for undermining the aims of the Bank and the International Monetary Fund. In response, the Japanese government set out to change the Bank’s core ideas about the role of the state in development strategy. It did so by inducing the Bank to pay more attention to East Asian development experience, so perhaps the Bank would change its mind, see more validity in the Japanese principles, and enhance Japan’s role as a leader in development thinking. Japan’s influence inside and outside the Bank would then grow. Specifically, the Japanese government persuaded the Bank to make a special study of East and Southeast Asia, focusing on why this region has become rich and what other countries should learn from the experience. The study was published in September 1993 as The East Asian Miracle: Economic Growth and Public Policy. In this paper we examine, first, the build-up of tension between Japan and the Bank; second, the process by which the study was written inside the Bank; and third, the resulting text. We shall ask whether Japan’s attempt to get the Bank to change its mind was successful. We shall see how the final document reflects an attempt at compromise between the well-established World Bank view and the newly-powerful Japanese view. The result is heavily weighted towards the Bank’s established position, and legitimizes the Bank’s continuing advice to low-income countries to follow the ‘market-friendly’ policies apparently vindicated by East Asia’s success. But the document also contains enough pro-industrial policy statements to allow the Japanese to claim a measure of success. Taken together with other Bank studies prompted by Japan at the same time, it provides a number of ‘attractor points’ for research and prescriptions more in line with Japanese views. Although the Bank emerges with its traditional paradigm largely unscathed, this particular episode may even be looked back on as an early landmark in the intellectual ascendancy, in East and Southeast Asia if not in the West, of Japanese views about the role of the state. Finally, we shall come back to the issue raised in the first paragraph—the autonomy of the World Bank, and the extent to which it can be regarded as an ‘actor’ with objectives and approaches that are not simply the vector of the interests of its member states. I. The World Bank’s Position in the Development Debate The World Bank enjoys a unique position as a generator of ideas about economic development. Around the world, debates on development issues tend to be framed in terms of ‘pro or anti’ World Bank positions. The Bank’s ability to frame the debate rests on, 1) its ability to influence the terms on which low-income countries gain access to international capital markets, 2) a research and policy-design budget far larger than that of any other development organization, and 3) its ability to attract global media coverage of its major reports. In the early 1980s the Bank swung into line with a us-led consensus about the needs of the world economy and appropriate economic policies for developing countries. Reflecting the demise of Keynesianism and the ascendancy of supply-side economics in the us and some parts of Europe, the consensus—the ‘Washington consensus’, as it has been called—was based on the twin ideas of the state as the provider of a regulatory framework for private-sector exchanges (but not as a director of those exchanges), and of the world economy as open to movements of goods, services, and capital, if not labour. The Bank’s new Structural Adjustment Loans applied conditions conforming to these ideas, such that borrowers had to shrink the state and open the economy to international transactions. Its annual World Development Reports have provided the conceptual framework and evidence to justify these conditions. In particular, the World Development Report 1987, entitled Trade and Industrialization, articulated a strong ‘free-market’ or neoliberal argument about the appropriate development approach.footnote3 The central problem of developing countries, in the Bank’s view, is the weakness of their ‘enabling environment’ for private-sector growth. The enabling environment consists of infrastructure, a well-educated work force, macroeconomic stability, free trade, and a regulatory framework favouring private-sector investment and competition. Policies to secure such an environment are collectively called ‘market-friendly’. The ‘market-friendly’ approach is not the same as laissez faire, the Bank is at pains to say, for there are areas where the market fails, in infrastructure and education, and where the government should step in with public spending.footnote4 On the other hand, the approach warns against intervention beyond these limits, especially against sectoral industrial policies designed to promote growth in some industries more than others. Market-friendly policies—neither complete laissez faire nor interventionism—are optimal for growth and income distribution, says the Bank. This set of ideas is broadly consistent with us demands that its trading partners—Japan in particular—change their domestic institutions in order to create a ‘level playing field’ for free and fair trade. In the late 1980s the Bank paid particular attention to financial sector reform. A Bank Task Force on Financial Sector Operations met to formulate policy on financial system reform, later to be put in the form of a mandatory Operational Directive. The Task Force championed a policy of far-reaching financial deregulation for developing countries, urging removal of all interest rate controls and all directed credit programmes. The World Development Report 1989, entitled Financial Systems and Development took a somewhat less extreme view. Written by a team that worked at the same time as the Task Force on Financial Sector Operations, it emphasized that private financial markets do make mistakes, particularly because of information problems and externalities—although these mistakes last for a shorter time than those of public financial agents. Where supervision and monitoring is effective, directed credit can work. Governments should, however, deregulate, but gradually. In August 1989, one month after the World Development Report 1989 was published, the Bank issued the Report of the Task Force on Financial Sector Operations—known as the Levy Report, after its chief author, Fred Levy. As noted, it took a strong view against government intervention in financial markets. The later Bank policy directive on financial sector operations took this report, not the World Development Report 1989, as its foundation. II. The Japanese Challenge Throughout the 1980s the Japanese state has hugely strengthened its external reach through aid programmes and foreign investment. By the early 1980s it was already the principal co-financier of World Bank loans, the number two shareholder in ida—the Bank’s soft loan facility—and the biggest source of bilateral aid for Asia. In 1984 it became the second biggest shareholder in the World Bank (ibrd) after the us. By 1989 it had the biggest bilateral aid programme in the world. In 1990 it became the second biggest shareholder in the International Finance Corporation—the Bank’s affiliate for private-sector lending. In 1992 it became the second biggest shareholder—equal to Germany—in the International Monetary Fund. By the early 1990s Japan passed the us to become the world’s biggest manufacturing economy; it accounted for half of the developed world’s total net savings—us savings accounted for 5 per cent; and it became the world’s biggest source of foreign investment. For all these reasons, Japan has come to matter for international financial institutions as never before—and also for the us state, whose deficits it has been financing. The Japanese government has encouraged its recipient governments—the us aside—to think more strategically and in more interventionist terms than can be accommodated by World Bank ideas. In particular, it has sanctioned attempts by low-income states to go beyond the conventional neoclassical tasks of providing a property-rights framework and moderating market failures due to public goods, externalities and monopolies. It has encouraged aid recipients to articulate national objectives and policy choices, to catalyze market agents, and to assist some industries more than others. The Japanese government claims that the potential benefits of the state’s directional thrust are illustrated by the actual benefits from the sectoral industrial policies of pre- and post-war Japan, and more recently, of Taiwan and South Korea. A regulated, ‘nonliberalized’ financial system capable of delivering concessional credit to priority uses, according to the Japanese, was a vital part of the organizational infrastructure of these policies. In line with this thinking, in 1987 the Ministry of International Trade and Industry (miti) published The New Asian Industries Development Plan, setting out a regional strategy of industrialization for Southeast Asian countries.footnote5 Responding to the appreciation of the Japanese yen in the mid-1980s and the resulting need to transfer more Japanese production offshore, the plan outlined the ways that Japanese firms making location decisions consistent with the plan would benefit from various kinds of aid for infrastructure, finance, market access, and so on. Officials were explicit that ‘Japan will increasingly use its aid... as seed money to attract Japanese manufacturers or other industrial concerns with an attractive investment environment’.footnote6 The Dispute over Directed Credit ‘Directed’ credit—meaning subsidized and targeted or earmarked credit—was to be a key instrument of this strategy. In the late 1980s Japan’s Ministry of Finance (mof) established the asean-Japan Development Fund, which offered directed credit to support private-sector development. The fund was administered by oecf, Japan’s largest aid agency. Unhappy at these developments, Bank officials expressed their reservations to Japanese officials informally—to no effect. In June 1989, a new Executive Director for Japan, Masaki Shiratori, arrived at the World Bank. As a senior mof official, he had helped steer Japan’s relations with international financial organizations for many years. Between 1981 and 1984, he played a central role in the strategy to raise Japan’s shareholding in the ibrd from number five to number two.footnote7 More persuasive in English than his predecessors, he was concerned to shift Japan’s role from cheque-writer to leader—‘no taxation without representation’, some Japanese comment wryly—and to make the Bank drop its blanket opposition to directed credit policies. By this time, both the World Bank and the Japanese government had well-articulated development strategies in place, the Bank emphasizing free markets, including nearly free financial markets, the Japanese government emphasizing guided markets, including guided financial markets. Japan was by then the second ranking shareholder in the World Bank after the us. And it had a new, articulate and forceful Executive Director, determined to make the Bank pay more attention to the East Asian experience and to rethink directed credit policies. In September 1989 the dispute between the Bank and Japan’s oecf over credit policies became explicit. Citing the case of the Philippines, a senior vice-president of the Bank wrote to the president of oecf—in charge of the asean-Japan Development Fund—asking him to reconsider the policy of subsidized targeted loans: passing these funds to the banks and final beneficiaries at below market interest rates ‘could have an adverse impact on development of the financial sector’ and hence ‘would create unnecessary distortions and set back the financial sector reforms’ which had been supported by the imf’s Extended Fund Facility and the Bank’s Financial Sector Adjustment Loan.footnote8 The dispute highlighted the underlying differences of view, the Japanese arguing that financial policies should be designed to advance a wider industrial strategy, the Bank insisting that credit should always and everywhere be at ‘market’ or non-subsidized rates.footnote9 Japan’s Executive Director made strong protests to the Bank’s senior management and to the Board of Executive Directors from member governments. Many Executive Directors from developing countries agreed with the Japanese position, but to no avail; Bank management refused to back down. Japan’s Ministry of Finance and its oecf began to fight back. A key figure was Isao Kubota, a senior mof official then on loan to oecf as managing director of the pivotal Coordination Department.footnote10 He did two things. First, he established a team to write a paper setting out the broad principles of the Japanese government’s understanding of structural adjustment. Second, he had discussions with Shiratori about how to get the Bank to pay more attention to the Japanese and wider East Asian development experience. This was the genesis of the Miracle study. Meanwhile tensions were growing between Japan and the us as well. From May 1989 through to 1992, the two states were negotiating over market access—the Structural Impediments Initiative. The us tried to make the Japanese undertake domestic reforms of such features as the retail distribution system and the cross-ownership of firms, so making it more like the ‘free market’ or American system. The Japanese mostly resisted and in turn urged reform of us institutions. An American business executive in Tokyo later said about the wider relationship between Japan, on the one hand, and the us and Europe, on the other: ‘The tired old technique of us and European leaders is to beat the Japanese with a piece of two by four. Not surprisingly, they resent it. They may be less cocky now that the economy is in recession, but there is a deep and growing and potentially damaging distrust of the West in the Tokyo corridors of power’.footnote11 Also during this period, Tokyo was flirting with membership in the Malaysian-sponsored East Asian Economic Caucus, from which the United States was excluded, while remaining cool to the American-endorsed Asia-Pacific Economic Cooperation forum.footnote12 This underlined Japan’s new willingness to pursue a course apart from, and even opposed to, that of the us. Back at the Bank, Lawrence Summers, a Harvard economist, joined as chief economist and vice-president in January 1991. Not noted for tact, he openly held the view that Japanese economists are ‘second rate’. From January to June 1991, drafts of the Bank’s World Development Report 1991: The Challenge of Development, underwent discussion within the Bank and the Board. Written under the leadership of a Chicago-trained economist, the report restated a largely free-market view of appropriate public policy for development, under the label ‘market-friendly’. The term was coined by Summers who exerted influence at this late stage of the report to moderate the extreme free-market position of the earlier drafts, but in Japanese eyes it still remained extreme. Blueprint for Development Then in October 1991 the oecf—whose main parent ministry is the Ministry of Finance—issued the paper initiated by Kubota, entitled ‘Issues Related to the World Bank’s Approach to a Structural Adjustment: Proposal from a Major Partner’.footnote13 Its main points are as follows: For a developing country to attain sustainable growth, the government must adopt ‘measures aiming “directly” at promoting investment’. These measures should be part of an explicit industrial strategy to promote the leading industries of the future. Directed and subsidized credit has a key role in promoting these industries because of extensive failures in developing countries’ financial markets. Decisions about ownership arrangements, including privatization, should relate to actual economic, political and social conditions in the country concerned, not to the universal desirability of privatizing public enterprises. For example, there may be legitimate national sentiments about the desirability of foreign ownership. ‘Japanese fiscal and monetary policies in the post-war era may be worthy of consideration. These were centred on preferential tax treatment and development finance institutions’ lendings.’footnote14 Also in October 1991, at the Annual Meeting of the Board of Governors of the World Bank and the imf, Yasushi Mieno, head of the Bank of Japan, the central bank,footnote15 said, ‘Experience in Asia has shown that although development strategies require a healthy respect for market mechanisms, the role of the government cannot be forgotten. I would like to see the World Bank and the imf take the lead in a wide-ranging study that would define the theoretical underpinnings of this approach and clarify the areas in which it can be successfully applied to other parts of the globe.’footnote16 Mieno’s statement was prepared by the International Finance Bureau of the mof. Isao Kubota—by now transferred back from oecf to a senior position in this same bureau, and drafter of Mieno’s statement—later made the point more vividly to reporters: ‘It’s really incredible. They think their economic framework is perfect. I think they’re wrong.’footnote17 By late 1991 tension between Japan and both the Bank and the us was running high. Articles based on interviews with Japanese officials began appearing in the American and Japanese press with titles like ‘Japan—us Clash Looms on World Bank Strategy’. The anonymous Japanese officials called the Bank’s approach ‘simple minded’, resting on ‘outmoded Western concepts that fail to take account of the successful strategy pursued by Japan and some of its Asian neighbours in developing their economies’.footnote18 Privately, those officials accused Bank economists of gross arrogance, of presuming to lecture them on why the Japanese government was doing the wrong things while at the same time asking for more Japanese money. Another statement of Japanese principles came out in April 1992, in the form of miti’s blueprint for economic reconstruction and development in Russia. ‘Western industrial countries’, it said, ‘can make many suggestions to help Russia with its economic reform. This paper...focuses on what Russia can learn from Japan’s experience...’ It described its approach as being in ‘stark contrast’ to that of the imf, presented in a report on Russia earlier the same year. ‘Market mechanisms cannot be almighty’, it claimed, expressing doubts about whether ‘macroeconomic approaches’, such as those advocated by the imf, were sufficient to meet the chief need of revitalizing production. Japan’s post-war economic renaissance could be used to formulate appropriate policies in, for example, the design of emergency measures to halt the plunge in output, and of ‘priority production programmes’ to ensure the supply of essential industrial goods. ‘The worst choice would be to diversify investment in an all-out manner, because...what is now most needed is focus on specific sectors of particular importance as a way to increase overall production.’footnote19 In other words, Russia must as a matter of urgency have a sector-specific industrial policy. III. Why the Japanese Challenge? As the Japanese government greatly increased its capital contribution to the Bank, it wished, not surprisingly, to see its views more fully reflected in Bank thinking. But why present itself as a champion of ‘antiparadigmatic’ views? As it becomes more powerful, why does it not endorse free trade and obscure the mercantilist elements in its own history? There are at least four possible reasons. The first is ideological conviction. The senior officials in those parts of the government leading the challenge genuinely believe that interventionist policies can be more effective than the Bank’s ‘market-friendly’ set of policies. They emphasize the role of interventionist policies in Japan’s own development—the ways in which selective interventions can help Japanese aid be more effective. Being able to demonstrate aid effectiveness is especially important when official development assistance (oda) has been largely exempt from the government’s budget cuts. With many Japanese policy-makers disgruntled about the amount of aid, the Finance Ministry is under constant pressure to show it being well used. Indeed, said mof officials, they have been criticized by other government agencies for being too focused upon aid effectiveness, for not paying enough attention to Japan’s national interest. Their reply is that making best use of aid money helps to stabilize the world economy, which is also in Japan’s national interest. Ideological conviction is especially intense on financial issues. The phrases, the ‘moneymaking culture’ and the ‘thing-making culture’, are in common use in Japan, representing a widespread sense—as in Islamic condemnations of usury—that making goods and providing services is intrinsically a more worthy activity than making money by financial dealings, and that the financial sector should be industry’s servant. The second reason for the challenge is organizational interest. The Bank’s criticism of Japan’s concessional and directed aid schemes in Southeast Asia were aimed at what mof considered its greatest post-war achievement. Directed credit was its principal industrial policy instrument in the post-war renaissance of Japan; effective use of directed credit is the foundation of its claim to have played a major role in the ‘miracle’. The claim is reflected in the oecf’s mandate to provide directed and concessional credit as part of the Japanese aid strategy. No wonder mof—and when we speak of ‘Japan—World Bank’ relations we mean mof—World Bank relations, for mof jealously guards its monopoly—resents hearing the Bank announce to the world that directed and concessional credit can never be effective, all the more so since the Bank’s claim rests on near total ignorance of directed credit in North-east Asia. The third reason is national material interest. Building a powerful market position across East and Southeast Asia is a top Japanese government objective. Interventionist policies can potentially help Japanese firms and the Japanese government consolidate profits and influence in the region—enabling the Malaysian government, say, to give special support to the Malaysian joint-venture partner of a Japanese firm, or to the Japanese firm directly through targeted loans and protection.footnote20 Getting the World Bank to admit the potential desirability of selective industrial promotion would help to advance this agenda. But why might the Japanese government, or a part of it, wish to advertise the fact that it was playing by different rules in its aid programme? The answer may lie in the fact that the Bank had already strongly criticized Japan for doing so. This put the burden on Japan to show that playing by different rules could yield development outcomes better than those of the Bank—or to get the Bank to rethink. The fourth reason is nationalism, the desire to overcome a sense of being judged inferior by representatives of other states—or in this case, multilateral financial institutions. This sentiment is caught in the phrase often heard in and about Japan, ‘economic superpower and political pygmy’, or in Ichiro Ozawa’s likening of Japan to a dinosaur with a huge body but a tiny brain.footnote21 In response to the perception of being judged inferior, Japan adopted a state strategy of channelling economic activities so as to achieve independence from, leverage over, and respect in the eyes of other states, rather than to achieve consumer utility, private wealth, or freedom of society from government. As it has become during the 1980s a ‘mature’ economy with a very large role in the international economy, it has also frequently been criticized for lacking the leadership on the world stage befitting its economic might. There is a growing urge among Japanese officials, politicians and the general public for Japan to set this right. But how? Japan cannot constantly bow to foreign—that is, us—pressure. It needs to be seen asserting its own views on appropriate rules for the international economy. These cannot be free-trade rules, for the free-trade ideology is already led by the us. It can differentiate its principles from those of the us by basing them on its own experience of economic nationalism, presenting them as general principles confirmed by other East Asian experience and as sources of meta-policies for developing countries today. On these grounds, it can present itself as the champion of developing countries in the governing councils of the international financial institutions. At the same time, its principles also stay away from the dangerous idea of Japanese uniqueness. No country has come to exercise a leadership role in the world system without claiming to represent a universalistic ideology. In short, the Japanese challenge to the World Bank can be seen as part of a wider attempt by the Japanese elite to develop an ideology that goes beyond Japanese uniqueness and yet remains distinct from free trade and orthodox liberalism.footnote22 IV. The Bank’s Resistance Bank managers saw the Japanese ideas about the role of the state—the emphasis on directed credit and the more general argument linking the appropriate role of the state to the amount of state ‘capacity’footnote23—as a serious threat. Why? First, because concessional credit from the Japanese aid budget makes World Bank credit less attractive. The Bank especially needs to find borrowers in East and Southeast Asia, where Japanese aid is concentrated, to raise the average quality of its loan portfolio. Second, the Japanese emphasis on directed credit as an instrument of the industrial policy of recipient governments runs flatly contrary to the Bank’s emphasis on financial system deregulation, a central thrust of its macroeconomic reform formula through the 1980s. Third, if the Bank were to embrace the interventionist role of the state wanted by the Japanese government it would, in the eyes of its managers, risk its ability to borrow at the best rates on world money markets—and so face lower demand for its now more expensive funds. It would also risk its second most valuable asset after its government guarantees—its reputation as a country-rating agency, a kind of international Standard and Poors that signals to private investors where they should put their money.footnote24 Why would such dire consequences follow? Because the Bank’s ability to borrow at the best rates and to act as a country-rat-ing agency depends on its reputation among financial capitalists, which in turn depends on its manifest commitment to their version of ‘sound’ public policies. Their version is based on the premise that only one set of rules should apply to all participants in the international economy and that those rules should express a non-nationalistic role of the state.footnote25 If this premise constitutes an imperative from the Bank’s point of view, it is because any change of mind could be very costly. Fourth, if the Bank were to embrace the Japanese view, it would run against the strategic and diplomatic power of the us, which has used the Bank as an instrument of its own external infrastructural power to a greater degree than any other state. And the Bank would delegitimize itself in the eyes of American academic economics, with its belief in the overwhelming virtues of markets and its political agenda of deregulation—an agenda endorsed by those who do well out of free markets. The President of the Bank has always been an American; Americans are greatly over-represented at professional levels in the Bank relative to the us’s shareholding; some two thirds of World Bank economists are certified by us universities—and 80 per cent by North American or British universities.footnote26 Fifth, the Bank’s constitution requires it to be ‘apolitical’, and the single meta-policy, sanctioned as it claims to be by a transcendent and apolitical ‘economic rationality’, helps the Bank preserve the claim of ‘political impartiality’. One of the most important conceptual contributions of the Bretton Woods conference—which created the World Bank and the imf—was the idea of equal treatment of all members of the new financial order. It was intended to avoid the politicization of the 1920s international rescue operations. There would be no ‘favourites’, but a community of states supporting each other at times of difficulty by means of a universalistic set of rules. To now admit the potential efficacy of sector-specific industrial policies would require the Bank to discriminate between countries in terms of such factors as government capacity and corruption, on the quite reasonable grounds that industrial policies are unlikely to be effective in states whose governments are thoroughly corrupt. But doing so would expose it to the charge of being ‘political’, and open it to pressure from borrowers saying, ‘You urged/allowed country X to do A, why can’t we?’footnote27 Sixth, commitment to the Bank’s meta-policy allows the organization to act quickly and concertedly. The meta-policy is derived from neoclassical economics and receives the endorsement of most us- and uk-trained economists who took control of the Bank from top to bottom over the 1980s; technical specialists—engineers, agronomists, health specialists, and so forth—were removed from operational management positions or not replaced when they retired.footnote28 The common commitment to the neoclassical meta-policy by the Bank’s management cadre helps senior management to overcome the ‘agency’ problem of subordinates exercising discretion in ways they do not like. It keeps the whole management spine in proper alignment. It also allows country departments to be efficient advice givers. Policies seen to be inconsistent with neoclassical normative theory are excluded from the start. Of course, the Bank’s lending practices on the ground have often differed from what the recipe calls for. But the case-by-case modifications come from the need to adjust pragmatically to ‘political realities’, not from a belief that the economics of the meta-policy might be less than universal. (So China, with one of the most interventionist, price-distorting governments of all, was the Bank’s fastest growing borrower over the 1980s. The Bank and China need each other—China to get finance and intellectual help, the Bank to lend to a big absorber with little debt.) At the level of principles, the neoclassical and largely free-market meta-policy is insulated from particular modifications. Seventh, the Bank sees the Japanese position as posing a threat not only to itself but to its borrowers. The Japanese position requires the low-income country state to play a strategic role in governing the integration with the world economy—maintaining the relative separation of the domestic and international spheres for policy making—not just the role of transmission belt from the ‘realities’ of the world economy to the national economy. Such a strategic role, says the Bank, generally lowers national welfare. Even if some evidence suggests that some governments some of the time have played this role effectively, 90 per cent of governments have been unable to. Notwithstanding this, the vested interests pushing governments to intervene in counter-productive ways are so powerful that governments will go on doing so unless hindered by some impartial and powerful agency—the World Bank and the imf, for example. The Japanese views, says the Bank, give unwelcome legitimacy to such interventionist impulses. Finally, even if such policies raise national welfare in a single case, they can do so only by ‘free riding’ on the restraint of others—promoting industries to compete in us markets while closing the domestic market to us exports, for example. So the Japanese principles cannot be practised by all at the same time, and in that sense pose a systemic threat. These eight reasons radically over-determine the Bank’s reaction of alarm and denial to the pro-interventionist views of its second biggest shareholder. But the danger could be diffused and confrontation contained as long as the Bank did not have to deal explicitly with the causes of East Asia’s economic success; in dealing with other regions or with ‘development-in-general’, it could simply ignore Japanese ideas about development strategy. On the other hand, if it did have to examine in depth the causes of this success, a more or less explicit statement about the validity of apparently very different views would have to be made. Given Japan’s power, that resolution would have to make some concession to Japanese views, for otherwise the number two shareholder would lose too much face and become less cooperative. The Japanese mof decided to force the issue. V. Making the Miracle: Stage One To recap: in 1989 the Bank made a strong criticism of the Japanese aid agency, oecf, for its credit policies in Southeast Asia. In response, senior mof officials considered how to get the Bank to be more ‘pragmatic’ and heed the experience of Japan and other East Asian economies. In 1991, soon after the arrival of Lewis Preston as the new president of the World Bank, the Japanese mof pressed the Bank to make a thorough study of East Asian development experience. The Bank’s senior management was reluctant to permit the study, but agreed for two reasons. First, the Japanese would pay for it, the Bank having to bear only the time cost of its own staff.footnote29 Second, in return for the Bank’s concession, the Japanese agreed to drop their opposition to the draft Operational Directive on Financial Sector Operations, which urged full-scale financial deregulation.footnote30 In January 1992 the study got under way, with a budget of $1.2 million from the Japanese trust fund. It was to be written over eighteen months for publication at the time of the Annual Meeting in September 1993.footnote31 The core study, giving the overall analysis and conclusions, was to be based in the Bank’s research complex under Lawrence Summers and Nancy Birdsall (the director of the research department, an American). They appointed John Page (DPhil in economics from Oxford, undergraduate in economics at Stanford, another American) to head the study. Page put together a team of six people, all with PhDs in economics, all but one from American universities.footnote32 None had adult experience of living and working in Asia. There were also to be a number of case-studies of countries organized by the Bank’s East Asia vice-presidency; some to be written by authors inside the Bank, others by outside consultants. The outsiders were offered $10,000 per case-study, and required to submit drafts in six months—so their research had to be largely off-the-shelf. In addition, several background papers on Japan were commissioned from Japanese scholars. Although it got the country studies, the East Asia vice-presidency felt passed over. The vice-president for East Asia, Gautam Kaji, first heard of the study at a board meeting. Asked by an executive director for his views about the proposed study, he confessed not to know about it. Summers bypassed the East Asia vice-presidency, aware that its senior managers and economists held views towards the free market extreme of the Bank’s range. The rivalry between the core team in the research complex and the East Asia vice-presidency was to shape the arguments of the study. At the same time, a parallel and complementary project was initiated, again with Japanese funding, to examine the effect of directed credit in Japan. This was undertaken on behalf of the Bank by the Japan Development Bank, reviewing its own programmes. Its conclusions were to feed into the Miracle study.footnote33 A third Japanese-funded study about Japan, ‘The Evolution, Character and Structure of the Japanese Civil Service, and its Role in Shaping the Interrelationships between the Government and the Private Sector’, was undertaken by the Bank’s educational arm, the Economic Development Institute (edi), for use in World Bank teaching courses. Suddenly the Bank was paying a lot more research attention to Japan than ever before, thanks to Japanese initiative and Japanese money.footnote34 From early 1992 to early 1993, the first drafts of the Miracle chapters were written and discussed within the core group. John Page was given a free hand by senior managers, with no hint of the expected conclusions. Lawrence Summers urged him to think in new ways, to listen carefully to the Japanese arguments. ‘We were eager to find a story that would be new, all the more so because the Bank’s standard “market-friendly” story had already been told in World Development Report 1991’, said Page later. Indeed, Summers’ reaction to Page’s proposed names for the team was: ‘Too neoclassical, you will be seen as trying to force East Asian data into a neoclassical strait-jacket’. Page responded that for the report to have an impact in the Bank, it had to use the language of neoclassical economics: the team stayed as he proposed. The team members accepted that East Asian governments implemented policies at substantial variance from the Bank’s orthodoxy, but they found it difficult to unearth clear evidence about the causal impact of these non-orthodox policies on economic growth. Wrestling with this issue for many months, they eventually concluded: ‘It is possible that some of these non-orthodox policies helped some of the time, but, with some exceptions, we can’t show it’. Also at this time, this version of the ‘institutional basis’ chapter was restructured. This version had taken as its main question, ‘What features of East Asian institutions enabled these economies to avoid the costs that befall equally interventionist and authoritarian states elsewhere; or why did their many strategic interventions not lead to massive rent-seeking?’ It presented government—business consultative councils, for example, as an institutional device that reduced the authoritarian character of East Asian political regimes by providing an institutionalized channel of feedback from the people directly affected by business policies. Birdsall and Page thought this might be interpreted as sanctioning authoritarianism and interventionism—as saying, ‘If you have institutional features X, Y, and Z you can avoid the expected costs of authoritarianism and interventionism’. In the rewriting, this theme was much diluted. The chapter was brought into line with the report’s larger argument that East Asian states are more successful because they are less interventionist, and the implication that some authoritarianisms are better than others was removed. VI. Making the Miracle: Stage Two Around March 1993 the second stage of the production process began with rounds of discussion at successively higher levels of the approval hierarchy. A full-time editor, Lawrence MacDonald, was hired from the Asia Wall St. Journal. footnote35 Over the next several months he and Page sent material back and forth, the editor revising the drafts in line with comments, Page commenting on the editor’s revisions, the editor taking on board Page’s comments and resubmitting to Page. The editor was the only person on the project with work experience in Asia. He attempted to inject some discussion of cultural propensities to save and educate, and of the role of the overseas Chinese. The team rejected these suggestions, the former for being too difficult to pin down with evidence, the second for being too liable to be taken as racist. To discuss the drafts, many meetings were held with people from the East Asia vice-presidency which had something close to a veto over the study being approved for sending to the Board—and thence for public release. The East Asia staff attacked the work for excessive emphasis on government intervention. ‘Where is the evidence for what you are saying?’, they demanded. The East Asia vice-presidency was well versed in demolishing arguments about the efficacy of industrial policy, its chief economist having just co-authored a book reiterating a largely free-market interpretation of East Asian economic success;footnote36 its vice-president, still smarting from being excluded from the study’s initiation, provided support for such challenges. Its representatives badgered the team about ‘strategy’—as in the working subtitle, ‘Strategies for Rapid Growth’ and phrases like ‘a strategic approach to growth’. Such phrases could be misconstrued to mean that East Asian growth was due primarily to ‘strategic’ interventions in industrial policy, or even to sanction the idea of an alternative East Asian type of capitalism. The East Asia representatives also argued, more generally, that the Bank had an interest in getting the market-friendly approach, as set out in the World Development Report 1991, accepted as the correct approach to economic policy in all developing countries, and it would look odd if the study of East Asia, of all regions, did not embrace it too. Not coincidentally, the World Development Report 1991 was written by a team headed by the man who was then chief economist for East Asia. The spectacle of the East Asia vice-presidency evacuating upon the draft convinced Page and Birdsall of the need to make concessions if the draft was to proceed up the approval hierarchy. What could they concede? First, they recognized that ‘strategy’ and ‘strategic’ implied—at least to the East Asia vice-presidency—a stronger argument about the efficacy of industrial policy than they wished to make, and were distracting attention from the substance of the argument about market failures. All references to strategy were therefore deleted, being replaced, where necessary, with the innocuous ‘functional’, as in ‘a functional approach to growth’.footnote37 Second, they praised the market-friendly approach in several places. Lewis Preston’s preface was made more explicit: ‘The authors conclude that rapid growth in each economy was primarily due to the application of a set of common, market-friendly economic policies’.footnote38 At this late stage, the editor was asked to write a box summarizing the ideas and evidence for the market-friendly approach.footnote39 He wrote, ‘In the past twenty years a consensus has emerged among economists on the best approach to economic development. . .These ideas have crystallized into what is now called the “marketfriendly” approach.’ By making these concessions, Page and Birdsall hoped to protect two key ideas in what they had earlier called ‘strategy’. One was that growth is a function of three sets of policies—those to foster accumulation, efficient allocation, and growth in productivity. Whereas the standard, market-friendly neoclassical argument stresses the need for good performance on all of four dimensions—macroeconomic stability, trade openness, human capital, and a rule-based system hospitable to the private sector—Page and Birdsall thought that there is some substitutability between the policies for accumulation, allocation, and productivity. Hence it is conceptually possible that costs in allocative efficiency (due to distorting industrial policies) are more than offset by gains in productivity (due to learning). The second idea is that markets—effective coordinating mechanisms for private agents in many contexts—may not work well for large and uneven investments in the early stages of development; for these, other mechanisms are needed, such as ‘deliberative councils’. But how to stop deliberative councils from becoming cosy havens for sharing out rents? Through contests between selected firms competing within tight rules and under the watchful eye of the government as referee.footnote40 Other parts of the report also came in for strong criticism from elsewhere in the Bank—all the more so now that Summers had leftfootnote41—but in the final version they appeared little changed. The section on directed credit and financial repression was attacked for making too many concessions to the view that these instruments could sometimes work. Page countered that the section did not, as the critics contended, repudiate the Bank line: it clearly stated that there is no proof that directed credit worked in Japan and Korea but also that the normal adverse effects of directed credit are not seen in those countries. Similarly on the wider question of financial repression, Page countered trenchant internal criticism by urging the critics to read carefully what the text actually said. While admitting the fact of financial repression in Japan and Korea, the text’s explanation for why the normal adverse effects on growth are not observed was not out-of-line with established Bank thinking: these effects are not apparent because the degree of repression has been moderate—thanks to macroeconomic balance and only slightly concessional interest rates for priority uses. This section of the report was of greatest interest to its Japanese sponsors. Its credibility was bolstered by the pre-eminent status in the American economics profession of its main author, Joseph Stiglitz, winner of the John Bates Clark medal for outstanding work by an economist under the age of forty. In the event, despite all the criticism, the section was left largely unchanged. The ‘institutional basis’ chapter, though already diluted, was attacked as the document proceeded up the hierarchy of approval. Many critics called for references to authoritarianism to be dropped. Birdsall and Page defended the chapter successfully, managing to retain oblique references to authoritarianism.footnote42 As the deadline loomed, intense effort was made to present a consistent message.footnote43 The Bank’s senior in-house editor was called in. He pasted each chapter page by page along the wall of a conference room. Together with several members of the team, he took a bird’s-eye view, suggesting how to bring the messages up front. He paid special attention to the headings, on the presumption that many readers do not go beyond them. Headings should themselves give the argument, he urged. Parts of the draft were revised to emphasize the neoclassical ‘fundamentals’. Results of the econometric tests of the effectiveness of selective industrial policy were rephrased to make them more clearly contra than in the original draft. Page later explained his principle for responding to criticism: if he agreed that the evidence was not strong enough to support a certain proposition, he toned down the statement, regardless of whether it was Bank orthodoxy or not. At the same time, he had to recognize that this was a World Bank document with an ‘anonymous’ author that sets out a ‘Bank’ position. So it should steer between the extremes, never straying outside the range of views represented within the Bank. Yet the team members were also anxious not simply to repeat the Bank’s standard line, and saw themselves as a vanguard pushing out the frontiers of debate. They were also well aware of the importance of Japan to the World Bank and of Japan’s interest in the conclusions. A senior manager later remarked: ‘Without the strong leadership of Larry Summers, Nancy Birdsall, and John Page, the report would not have moved anything like as far [from Bank orthodoxy] as it did.’ VII. Argument and Evidence The final document bears traces of the three-way tussle between Japan, the research vice-presidency, and the East Asian vice-presidency. It concedes for the first time in a major Bank publication the fact of extensive government intervention in most of East Asia. It also grants the argument that some of these interventions, in the areas of exports and credit, may have fostered growth and equity in some parts of East Asia. Further, the report states that ‘More selective interventions—forced savings, tax policies to promote (sometimes very specific) investments, sharing risk, restricting capital outflow, and repressing interest rates also appear to have succeeded in some hpaes [High Performing Asian Economies], especially Japan, Korea, Singapore, and Taiwan, China.’footnote44 And again: ‘Our evidence leads us to conclude that credit programmes directed at exports yielded high social returns and, in the cases of Japan and Korea, other directed-credit programmes also may have increased investment and generated important spill-overs.’footnote45 Lewis Preston’s preface is significant because it is the President who ultimately must keep the main shareholders happy, and in this case the number two shareholder evidently needed to be made less unhappy. The preface was written within the core team and did not have to fight its way past the East Asian vice-presidency. It says, for example, ‘This diversity of [East Asian] experience reinforces the view that economic policies and policy advice must be country-specific, if they are to be effective. . .The report also breaks some new ground. It concludes that in some economies, mainly those in Northeast Asia, some selective interventions contributed to growth, and it advances our understanding of the conditions required for interventions to succeed. . .These prerequisites suggest that the institutional context within which policies are implemented is as important to their success or failure as the policies themselves.’footnote46 These are Japanese-style statements. Despite all the pressures for the Bank not to admit it has been wrong, the President of the Bank here hints at just that. The preface does not even use the normal protective cover; it says ‘some selective interventions contributed to growth’, without the ‘may have’. A cynic might say that the ‘some selective interventions contributed to growth’ statement by Preston, plus the line in the text on page 356 (‘other directed-credit programmes also may have increased investment and generated important spill-overs’) are the nuggets for which the Japanese paid $1.2 million. The rest of the text takes a much stronger anti-industrial policy line. The flavour of the overall document is expressed in statements like ‘industrial policies were largely ineffective’, and ‘We conclude that promotion of specific industries generally did not work and therefore holds little promise for other developing economies’.footnote47 It is not surprising that the bulk of the report gives a strong endorsement of established World Bank ideas. We saw earlier why the Japanese ideas constitute a serious threat. But the Bank cannot credibly reject ideas just because they are a threat. It has to claim to reject them on the evidence—of which the Miracle provides lots for its anti-industrial policy arguments. The trouble, as several analysts have shown, is that most of the evidence does not survive serious scrutiny.footnote48 Here are three examples: 1) The key proposition that more open economies grow faster than closed ones is based on the finding that indicators of openness are positively correlated with growth in the basic growth regression. One indicator of openness is an index constructed by David Dollar. As Dani Rodrik argues, the index is really a measure of real exchange rate divergence, not of openness.footnote49 But if used as an index of openness, Dollar’s own published results reveal that Japan and Taiwan were less open during 1976–85 than Argentina, Brazil, India, Mexico, the Philippines, and Turkey—a result ignored by the Miracle study. Rodrik concludes that the evidence presented for the proposition that more open economies grow faster is simply not relevant. To the extent that it is, it points the other way. And here as throughout, had China been included, the evidence would have pointed still more strongly the other way.footnote50 Since the early 1980s, China has been outperforming most developing countries, yet it has remained—while liberalizing—much less liberal than most, with extensive controls on finance, trade and industry. 2) The report says that ‘price distortions were mild’, or that ‘East Asia’s relative prices of traded goods were closer on average to international prices than other developing areas’.footnote51 This generalization is important for the argument that, while industrial policies existed in East Asia, their magnitude was slight. But the report also acknowledges that the relative prices of Japan, Korea, and Taiwan deviated more from international prices than those of such notorious interventionists as India, Pakistan, Brazil, Mexico, and Venezuela in 1976-85, another finding it does not comment upon.footnote52 How does it reach the vital conclusion about low average price distortions? By averaging the price distortion scores of all eight East Asian cases, including the Hong Kong and Singapore minnows. 3) One of the tests of the effectiveness or otherwise of sector-specific industrial promotion uses the correlation between growth in output or value added by different industries, and the level of wages or value added per worker in the same industries.footnote53 If sectoral industrial policies made a difference, the argument goes, we expect a positive correlation, because industrial policies aim to favour capital- and technology-intensive industries and these factor intensities are proxied by high wages. So if industries that grow faster also have higher wages, this means that the more capital- or technology-intensive industries are growing faster, and industrial policies can be declared successful. Conversely, if the correlation is negative we have grounds for concluding that structural change is driven not by industrial policy but by market forces. It can be argued that the test is mis-specified.footnote54 But the problem is with what the report does with its own evidence. The results for several time periods yield mostly positive correlations (pro-industrial policy) for Hong Kong, and Japan, and mostly negative ones (pro-free-market forces) for Taiwan and Korea. But none of the results is statistically significant—except the negative correlations for Korea. The report still concludes that these results confirm the ineffectiveness of industrial policy in East Asia. The Middle Road Once such standards of inference are allowed to leak into what we call ‘evidence’, confirming results can be pumped out like bilge water.footnote55 It is a fine irony that when the one member of the team with work experience in Asia suggested some discussion of cultural propensities to save and educate, he was told the matter could not be discussed because of lack of evidence. The weakness of evidence notwithstanding, the argument sweeps to its paradigm-protecting conclusions on the strength of several rhetorical techniques. One is to structure an argument as a triptych with two extremes and a middle, our confidence in the middle being elevated by the foolishness of what flanks it. In the Miracle we are shown two cartoonish interpretations of East Asian success—laissez faire and government intervention—and then the sensible market-friendly approach in between. This was, however, a late addition. Together with the removal of ‘strategy’and ‘strategic’, it was part of the price of acquiescence from the East Asian vice-presidency, and the means by which the chief economist of the East Asia region could propel the conclusions of his World Development Report 1991 to the forefront of the Bank’s thinking on development. The report also seeks to persuade by ignoring serious alternative explanations of East Asian economic success. The main alternatives to such ideas as ‘market-friendly policies plus export-push policies yield export-led growth’ are not ‘laissez faire’ or ‘government intervention’. Indeed, no serious scholar has argued that the difference between East Asia and elsewhere is to be explained mainly in terms of government intervention. The main alternative, rather, is ‘favourable initial conditions—especially human capital and infrastructure—plus investment-led growth’. The causality runs from higher investment to faster technical change and higher imports, and from these to higher exports—these exports being more a result than a cause.footnote56 Certainly, export growth helped to maintain the key driving force—high rates of return on accumulation (by permitting economies of scale), but so, too, did rising skill levels and an array of government policies designed to boost productivity and keep the lid on income inequality. Sectoral industrial policies enter the explanation as an important cause of high rates of aggregate investment as well as a cause of the structure of that investment, helping East Asia to move quickly from the ‘factor-cost’ driven stage of competitiveness to the ‘investment’ driven stage.footnote57 Of course the report notes the fact of unusually high investment in East Asia, but sees it as more a result of market-friendly policies and export-push than as being itself the primary proximate driver—though without doing the econometric tests to examine the causality. As for the fast growth of Southeast Asia—Thailand, Malaysia, Indonesia—the report assumes the causes to lie in domestic factors, and fails to examine the extent to which their growth can be explained in terms of spill-over effects from the fast growth of the more nationally focused, governed-market economies of East Asia. Furthermore, the report tries to persuade by employing asymmetrical standards of evidence. As the drafts progressed, the many critics who asked ‘what exactly is your evidence?’ were concerned only with the pro-intervention propositions. They took for granted that if the evidence was not compelling it should be discounted, but did not apply the same scrutiny to propositions in favour of the free market. The market is innocent until proven guilty, the government is guilty until proven innocent. Finally, the report fails to make explicit some key distinctions, with the effect of allowing readers more scope for interpreting the results in line with their preconceptions. The striking case in point concerns credit. The Japanese were especially interested in getting the Bank to admit that directed credit—targeted at particular sectors—had worked in Japan and elsewhere in East Asia. But the Bank is deeply committed to the view that selective industrial promotion cannot raise national welfare, and so needs to conclude that it did not do so in East Asia. Since directed credit is, it would seem, simply one instrument of selective industrial policy, the two propositions—Japan’s directed credit, the Bank’s selective industrial policy—cannot both be true at the same time. Yet the report manages to imply that they are. It does so by classifying interventions into three ostensibly non-overlapping categories: selective industrial policies, directed credit policies, and export-push policies. It concludes that the first failed, the third worked, and as for the second—the focus of Japanese interest—it states, as we have seen, ‘that credit programmes directed at exports yielded high social returns, and, in the cases of Japan and Korea, other directed credit programmes also may have increased investment and generated important spill-overs’.footnote58 On the face of it, this says that for Japan and Korea directed credit may have been effective as an instrument of sectoral industrial policy, though the report also claims that sectoral industrial policy did not work. Dani Rodrik writes that ‘It is difficult to fathom how [such a logical inconsistency] found its way into the report (and as a major conclusion, to boot).’footnote59 Part of the reason was an editorial failure to make a clear distinction between two types of ‘directed’ or ‘selective’ policies: ‘functional’ and ‘sectoral’, where ‘functional’ refers to a non-sector-specific function, like r&d or exports, and ‘sectoral’ refers to specific sectors—chemicals or machine tools, for example. When the text talks of ‘selective’ industrial policy it means ‘sectoral’ or ‘sector-specific’; when it talks of ‘selective’ or ‘directed’ credit policy, however, it means ‘functional’. Its only evidence on directed credit for other than exports comes from a study of the effects of subsidized r&d credit in Japan—that is, a study of a functionally-directed, not sectorally-directed credit policy. On the basis of this study, the report says that (functionally) directed credit worked in Japan in the sense that it had higher social returns than private returns, made a net addition to r&d investment rather than substituting for more expensive commercial credit, and was cut off when no longer needed. So the Bank’s conclusion about directed or selective credit applies to functionally-selective policy, while its conclusion about selective industrial policy applies to sectorally-selective policy. Why was such an obvious source of confusion allowed to persist? The effect of fudging the distinction between functional and sectoral was to allow those sympathetic to the Japanese position on credit to infer a greater agreement with that position than was actually the case. VIII. Responses In August 1993 the World Bank executive directors (eds, the represen tatives of member states who act as overseers) considered the final draft. Their reactions showed nothing like consensus. The us ed gave a glowing endorsement of what he took to be the free-market message of the report. (Some of the core team were disturbed to hear how he spin-doc-tored all their qualifications away.) The newly arrived Japanese ed was cautiously complimentary. The Argentinian ed said, angrily, that the whole report was an apologia for interventionism. The Indian ed came close to saying that the report’s anti-interventionist conclusions were fixed in advance and the evidence tailored to fit. Few changes were made in response to the Board’s comments. If one is being attacked from all sides, Page later explains, the argument must be about right. Indeed, unknown to the eds, the document had already been typeset by the time of the Board discussion to ensure readiness for the Annual Meetings. It could not have been changed even if the eds agreed on changes.footnote60 The incident illustrates the independence of the Bank staff from the Board, despite the Board’s status as the supervisory body representing member countries. On September 26, 1993, exactly on time, The East Asia Miracle: Economic Growth and Public Policy was launched at the Annual Meetings of the World Bank and imf. There was a press conference, a press release, and a seminar for Annual Meeting participants. The report ‘sells itself’, because of outside interest. The diversity of views among the eds was a microcosm of reactions outside the Bank. In the press, for example, some journalists (mainly Japanese) said that the study confirmed the effectiveness and replicability of East Asia’s government interventions. Others (mainly American and British) said that it confirmed their ineffectiveness and unreplicability. The London Financial Times led its review of the report with, ‘Industrial policies to promote particular sectors or companies have been a failure in East Asia and do not explain the region’s rapid growth in recent decades, according to a World Bank study’. The Nihon Keizai Shimbun, Japan’s leading business paper, said ‘the report cites the accumulation of high-grade human and physical capital as a motivating force and highly evaluates the effects of government intervention . . .’footnote61 mof officials celebrated the fact that the Bank had at last admitted that state intervention can be useful, but were also critical of some of the conclusions. In December 1993 former executive director Masaki Shiratori, now posted to the oecf, delivered a hard-hitting critique at a seminar in Tokyo. He argued that ‘Comparative advantage should be regarded as a dynamic notion rather a static one. . .It is theoretically right to pick and nurture specific promising industries which do not have comparative advantage now. Many developing countries desperately need to get rid of the monoculture in such commodities as coffee, cocoa, copper and tin, which resulted from static comparative advantage. .A latecomer to industrialization can not afford to leave everything to the market mechanism. The trial and error inherent in market-driven industrialization is too risky and expensive considering the scarcity of resources.’ He went on to make a number of theoretical and technical points against the Total Factor Productivity test of the effectiveness of selective industrial policy, concluding, ‘In view of these theoretical and technical problems in the Report’s analysis of industrial policy, I hope further studies will be made within and outside the World Bank. In the meantime, I sincerely wish that the Bank will adopt “pragmatic flexibility” in prescribing policy advice to developing countries’.footnote62 Isao Kubota concluded his remarks at the same seminar, saying: ‘Perhaps the best lesson could be that policy makers and policy advisers, including those in the World Bank, should not be dogmatic but be pragmatic. For that purpose modesty, not arrogance, and a sincere attitude toward finding the right policy measures, are essential’.footnote63 A senior mof official close to the Miracle study characterized mof and miti reactions as follows: ‘mof people consider this a good step forward, although they are not fully satisfied with the study’s negative assessment on industrial policy. The reaction of the miti people is mixed: they share the mof view, on the one hand, but they are afraid to be accused of excessive intervention now in the course of negotiations with the us and the ec’. He referred to miti concerns as expressed by, for example, Makoto Kuroda, miti’s best known hard-line negotiator with the us: ‘We must not provide a dangerous basis for the argument that says Japan conducts itself by a different set of rules and must be treated differently. . .For some time I have repeatedly stated that we should avoid expressions such as “Japanese-style practices”.’footnote64 Opinion about the Miracle study within miti differed between the two key bureaux, the International Trade Bureau and the Industrial Policy Bureau. The former is preoccupied with maintaining access to the American market, for which avowed commitment to ‘free-market’ and ‘level playing field’ symbols is important; people from this bureau tended to be enthusiastic about the study’s conclusion that selective industrial policy has, by and large, been ineffective in East Asia. The Industrial Policy Bureau, by contrast, is committed to boosting the idea of miti’s successful steerage of the Japanese economy, and people from this bureau tended to be more critical of the study. mof’s critical stance may reflect its concerns to maintain a strategic aid programme using directed credit and other infant industry incentives. The two agendas—that of the International Trade Bureau of miti and that of mof—may reflect a single higher-level strategy: to maintain access to the American market over the five-year middle-run, while building up a dense presence in the Southeast Asian and China markets for the ten-year longer run, at the end of which these markets are expected to be more important than the American.footnote65 Within the Board of the Bank, Shiratori’s successor as Japan’s executive director was less active. He did not push the concerns that lay behind Japan’s promotion of the Miracle study. This may reflect a high-level decision in Tokyo to calm relations with the World Bank in order to avoid causing even more turbulence in Japan’s relations with the us. As part of this calming strategy, the Japanese government agreed with the Bank that Japan’s directed credit programmes, though they continue, will not use narrow earmarking (will not define beneficiaries narrowly) and will not have a big subsidy element (not more than one or two percentage points below the market rate). As for the Bank’s response, a top manager said ‘We simply cannot afford to take a more custom-tailored approach to lending conditions, as the Japanese have been urging. If we were to say to the Philippines, “It is ok for Malaysia to do this but not for you”, we could be accused of violating the political impartiality condition of our charter.’footnote66 No follow-up research has been planned. The director of the research department explained that ‘the real issue is the relevance of the East Asian experience for other developing countries. . .Now the East Asian study is completed, the research agenda lies more in Africa and other developing countries than it does in East Asia.’footnote67 He took for granted that ‘the East Asian experience’ is the experience as interpreted in the East Asian Miracle. IX. The Art of Paradigm Maintenance—and Change Our story raises a more general question. How does the World Bank—a large institution, with some four thousand professional staff drawn from many countries,footnote68 producing dozens of public reports a year—man-age to deliver what the outside world hears as a single central message? The art of paradigm maintenance begins with the choice of staff. As noted, about 80 per cent of Bank economists are North American or British trained, and all but a few share the preconceptions of mainstream Anglo-American economics.footnote69 If they were to show sympathy for other ideas—if they were to argue that sectoral industrial policies can in some circumstances be effective, for example—they would be unlikely to be selected for the Bank, on grounds of incompetence. The organization’s few non-economist social scientists are employed on marginal issues like resettlement and participation, like anthropologists by colonial administrations before them. But within the staff there remains a range of views that command some following. The second technique of paradigm maintenance is the internal review process. A document goes through rounds of discussions at successively higher levels of the hierarchy, each level being a filter that narrows the range of views espoused by ‘the Bank’. It is not just that higher levels are more concerned with the Bank’s and the system’s integrity than with the integrity of the research. It is also that promotion criteria select people for the higher levels who make decisions quickly and with closure, using ‘facts’ selectively to support pre-conceived patterns and convictions. Such people tend to be intolerant of those who do not share the conclusions to which they leap.footnote70 Thirdly, the legions of Bank editors, some in-house, some employed as consultants, are a part of the maintenance mechanism. Their continued employment depends not only upon their ability to write clear English but also on their ability to write copy that, being in line with ‘Bank thinking’, will not attract criticism. This is the review and editing mechanism. The criteria applied are partly formal and partly substantive. The formal criteria relate particularly to the need for a clear ‘message’. Great emphasis is placed on having a clear message, on minimizing ‘on the one hand, on the other hand’ statements which are thought confusing to the intended audience of policy-makers. (Indeed, the early meetings of the writing team are often taken up with discussion of ‘What are going to be the key messages of this report?’, before the research is done.) The message is to consist not of a setting out of possible alternatives and conditions in which they make more or less sense, and still less of acknowledgement that the evidence is mixed or insufficient, but is to consist of the best policy for the ‘typical’ developing country. This makes for ‘clarity’. The need for a clear and consistent message for policy-makers has implications for the content of the message. The members of the team, partly propelled by professional norms, may be concerned to speak the truth as they see it. But at the higher levels reviewers are sensitive to the more ‘systemic’ pressures for paradigm maintenance discussed earlier—the need not to upset capital markets, and the self-perception of the Bank as a bulwark against the vested interests that push governments to intervene in socially counter-productive ways. Their comments page by page are unlikely to allude directly to these systemic pressures. Rather, they insist that everything should fit the overall message. This is the mechanism for conformity. All prominent Bank documents go through it. But what issues get onto the Bank’s agenda in the first place? On the whole, the Bank has been a reactive rather than proactive organization, taking its lead from outside. The Bank ensures its own expansion and centrality by launching bids for expert status on some of the issues at the top of the current agenda of development debate, proposing market solutions with compensatory or mitigating elements, creating a consensus around its position, and marginalizing more radical alternatives.footnote71 Outside the Bank, the debate then tends to configure itself into ‘pro- or anti-’ Bank positions. This might be called, tongue-in-cheek, a Strategy for the Sustainable Development of the World Bank. The East Asian Miracle can be read as the latest expression of this strategy. East Asia and industrial policy came to centre stage in the late 1980s, as the us and European economies continued to limp and East Asian economies continued to soar. The new element in the situation, compared to, say, a report on Africa or the Bank’s poverty work in the late 1980s, is that the number two shareholder was putting pressure on the Bank to endorse, or at least make some concession to, its non-orthodox views about development principles. The mere centrality of the issue in the development debate would not have been sufficient to prompt the Bank to make a special study, for the issue was at once too indirectly tied to lending and too likely to annoy the Japanese or to complicate the Bank’s policy formula.footnote72 But when Japan agreed to pay for the study and to drop its opposition to the operational directive on financial sector reform the Bank could not say no. These initiating circumstances made it important for the team leader to be someone known to be solidly in the mainstream of Bank thinking, not a doctrinaire free marketeer. John Page met this condition; his pedigree, as a student of Ian Little’s and protégé of Anne Krueger’s,footnote73 was conservative, but he had subsequently espoused more pragmatic views. Likely candidates from the East Asia vice-presidency were either free marketeers or too much under their hierarchical command. Even so, the universalistic and non-institutional ethos of neoclassical economics meant that no premium was given to selecting people for the core team who had expertise in East Asia—whether Bank staff or consultants. Any Bank economist is expected to be an expert on a country or region within a matter of months. As we have seen, the East Asia vice-presidency was excluded. True, it got the country studies, but these were largely ignored by the core team. Yet the East Asia vice-presidency could not be prevented from being the major reviewer, because in the higher level review committees the East Asia vice-president met the Research vice-president on equal terms—and with much more personal influence in the Bank where he had spent his whole career. If the East Asia vice-president decided to do so, he could effectively prevent or at least delay the report in its path to the Board, and so hinder its publication. The cross-pressures among the Japanese sponsors, the core team, and the East Asian vice-presidency help to explain the report’s inconsistencies. Inconsistency as a Register for Change The inconsistencies should not be seen simply as ‘mistakes’. The authors may have left them in—to the extent that they were aware of themfootnote74—in an attempt to widen the grounds of debate without generating a backlash that would cause the report to be dismissed as incompetent or ideological, and the Bank to be accused of changing its mind. The pro-industrial policy statements, though at odds with the rest of the report, may function as attractor points by enabling those wishing to put new questions on the agenda to claim legitimacy from the Miracle study. This, it could be argued, is the most likely way that big organizations change their minds; sharp changes are rare. The Japanese have influenced the Bank enough to provide attractor points beyond those in the Miracle study itself. The several studies of Japanese economic policy and civil service organization sponsored by the Bank at about the same time—and also paid for by the Japanese—provide a set of policy ideas that can legitimize further work in these domains, outside and inside the Bank. In particular, the Bank’s impri matur can help legitimize the idea of ‘Japan as model’ for Japan’s use in its own more dirigist Asian aid strategy, further strengthening the constituency for these ideas. It may also be argued that the Bank’s softening of its stand against directed credit, as of 1995, owes something to the wider Japanese pressure on the Bank. Compared to the 1980s, the Bank is now less likely to insist that directed credit and interest rate subsidies should always be avoided. It is more likely to insist simply that the onus must be on the proposer to explain the special circumstances justifying directed credit in a given case.footnote75 The shift is small but not trivial, and gives the Bank more flexibility in responding to Japan’s continued use of directed credit. Although the Japanese government has ceased pressuring the Bank, it has not stopped promulgating its ideas in developing countries. Seeing ‘the Japanese approach to industrial policy’ as a new export product, it is building up an enormous capacity for teaching Asian bureaucrats, industrialists and scholars about the Japanese approach to industrial policy. One of the leading figures in this campaign recently declared, ‘Free market theory has failed in many areas like Russia, Eastern Europe, and Sub-Saharan Africa because it is too short sighted and too market oriented. Not enough attention was paid to these countries’ own economic and social structures. . .Japan started from a planned economy post war, to become gradually liberalized over the years. I would say we are now 80 per cent of the way to being a free market economy. In developing countries it should be more like 50 per cent. We are not saying that developing countries should imitate Japan. But they do need to study an alternative to neo-classical economic theory’. To supply them with such an alternative, in 1995 between 500 and 600 foreign government officials will attend courses in economic development run by the ministries of international trade and industry, finance, foreign affairs, and the Bank of Japan. Scores of Japanese officials will also leave Tokyo on secondment to governments in developing countries, or to swell the small ranks of Japanese officials in multilateral development agencies. Most of the countries targeted for receiving this attention are also lucrative markets for Japanese goods.footnote76 The argument raises two wider points. The first is about the Bank’s research function. The Bank’s legitimacy depends upon the authority of its views; like the Vatican, and for similar reasons, it cannot afford to admit fallibility. At the same time, many of the Bank’s research publications, especially the high visibility ones like the World Development Reports, are really advocacy statements, steered by the bedrock perception that the Bank must act as a counterweight to all the gravitational pulls towards excessive government intervention—which justifies erring on the side of markets. Hence for good organizational and political reasons the Bank’s research is biased towards the conclusion that ‘there is no alternative’ to government policies that stay within the bounds of ‘strengthening the enabling environment for private sector development’. The Bank’s endorsement of this tenet is important for its authoritative image in the eyes of the interlocking social groups who embrace the ‘Washington consensus’. The research must also be largely quantitative, for numbers and econometric technique themselves confer authority. Research that meets these criteria thus helps to maximize staff commitment internally and authoritative reputation externally, and in turn colours the ‘reality’ against which those leaders of economic opinion check their expectations of the future. But its conclusions are not necessarily those that are most consistent with the evidence. The second point concerns the Bank’s autonomy. Our case study shows the Bank fending off a challenge to its way of seeing from its second largest shareholder. On the face of it, this looks like autonomy. It seems consistent with William Ascher’s argument that ‘the viability of a development objective or strategy to be implemented through the World Bank depends not only on the acquiescence of the obvious international actors—the nation states through their formal institutional representation and their various pressures—but also on its congruence with the professional role models of the relevant staff. If the staff perceives the strategy or objectives as a “decline in standards”, as requiring them to become more “political” vis-à-vis the borrower governments. . .its viability is doubtful unless altered role models can be quickly inculcated, new incentives provided, or rapid staff turnover undertaken.’footnote77 The problem is not that this argument is wrong, as far as it goes, but that it stops short of asking about the structure of power in which the Bank operates, and how that structure affects the Bank’s response to new development approaches. The story of The East Asian Miracle shows the determining importance of essentially American values and interests in the functioning of the Bank.footnote78 But the influence is exerted not mainly from the American government to the senior management of the Bank—if we look just at this relationship we see considerable autonomy, though the President has always been American. The influence comes partly through the Bank’s dependence on world financial markets, and the self-reinforcing congruence between the values of the owners and managers of financial capital and those of the us state. It also comes through the Bank’s staffing and professional norms. Not only are Americans greatly over-represented in the professional and managerial ranks but, at least as important since the beginning of the 1980s, is a second channel of influence—the conquest of managerial positions by economists, and the recruitment of economists, including some from the developing countries, predominantly from North American and British universities (virtually none from Japanese universities). This channel of influence is obscured by talking of ‘professionalism’ as a source of the Bank’s autonomy, without also talking about the content of that professionalism and from which member state’s intellectual culture it comes. By examining such factors we can see how the Bank forms part of the external infrastructural power of the us state, even though it by no means bows to every demand of the us government. Whereas the Japanese state uses its strong domestic infrastructural power directly to leverage its external reach—especially in Southeast Asia and China—the us state, with much weaker domestic infrastructural power, relies upon its dominance of international organizations like the World Bank and the imf to keep those organizations pursuing goals that augment its own external reach. The Bank’s stance as honest broker allows it to insist on the acceptance of those goals more openly than the us could itself. The story of The East Asian Miracle shows how this process worked itself out in one particular case.